The Ecological Associations of Surface-Dwelling in Qom Province in the Northwest of Central Plateau of

Mehregan Ebrahimi1*, Faraham Ahmadzadeh2,3, Hossein Mostafavi2, Aahmad Reza Mehrabian4, Asghar Abdoli2, Abdolrasoul Salman Mahini5

1 School of Biological Sciences, Flinders University, Adelaide, South Australia, Australia, 2 Department of Biodiversity and Ecosystem Management, Environmental Sciences Research Institute (ESRI), Shehid Beheshti University, Tehran, Iran, 3 Zoologisches Forschungsmuseum Alexander Koenig, Adenauerallee, Bonn, Germany, 4 Department of Botany, Faculty of Biological Sciences, Shahid Beheshti University, Tehran, Iran, 5 Department of the Environmental Sciences, Gorgan University of Agricultural Sciences and Natural Resources, Gorgan, Iran

Abstract

We used pitfall trapping to investigate the effects of elevation, plant density and soil structure on diversity and the impact of these habitat factors on habitat selectivity in the Qom Province in the Central Plateau of Iran. From a total of 12 1-ha plots, we captured 363 individuals of 15 species of lizards (six species of , five species of Agamidae, two species of Gekkonidae, one species of Varanidae and one species of Scincidae). A generalized linear model (GLM) determined that elevation was the most important factor impacting species diversity. The highest species diversity was at the intermediate elevation (1289 m). Abundance of 6 out of 15 species showed strong relationships with some habitat factors. These relationships were demonstrated by habitat selectivity index (Ivlev's index). Our result supports other surveys that showed that elevation plays an important role in determining lizard species diversity.

Citation: Ebrahimi M, Ahmadzadeh F, Mostafavi H, Mehrabian AR, Abdoli A, et al. (2013) The Ecological Associations of Surface-Dwelling Lizards in Qom Province in the Northwest of Central Plateau of Iran. PLoS ONE 8(12): e83890. doi:10.1371/journal.pone.0083890 Editor: Ulrich Joger, State Natural History Museum, Germany Received October 15, 2012; Accepted November 18, 2013; Published December 13, 2013 Copyright: © 2013 Ebrahimi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: The authors have no support or funding to report. Competing interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction richness and evenness) [9,10], and biomass of lizards can vary among habitats, because each species responds differently to Ecologists have investigated patterns of species diversity abiotic and biotic factors. Variation in topography provides and the mechanisms of community assembly for many years transitions across multiple habitats on a regional scale [1], and lizards have often been used as model systems to (mesoscale), which may show the role of elevation in explain patterns of species diversity [2-4]. Lizards show community structure [11]. Species diversity, across all taxa, extensive inter- and intraspecific variation in patterns of space generally decreases with increasing elevation [7], but this use [3,5] and distribution, depending on environmental factors decline is not necessarily linear, which has been shown in such as elevation, soil structure, vegetation [6,7] , as well as some species of birds and insects [12]. Amphibians and temperature and precipitation, which are usually correlated with follow this general pattern and their species diversity elevation. These physical properties of the environment have decreases with increasing elevation [13]. an important role in defining the relationship between habitat More recent studies have shown that there could be three and lizard species diversity. For instance Shenbrot and other patterns of species diversity. All three patterns Krasnov [8] classified habitat types using factors that described demonstrate a generally decreasing species diversity with vegetation and soil structure. Then they reported that lizard increasing elevation but with a diversity peak at low plateau, richness, diversity and biomass changed among these habitat mid-elevation or at both [14]. McCain [14] also showed that types within different landscapes. For instance they showed ambient temperature is strongly correlated with these patterns olivieri avoids habitat with soil structure that have (low plateau, mid-elevation and low plateau with mid-elevation), moderate gravel content. in particular along wet mountain gradients. In these areas The species richness (actual number of species), diversity opportunities to use radiant heat for thermoregulation are (effective number of species, which takes into account both limited. On the other hand, in arid environments the effect of

PLOS ONE | www.plosone.org 1 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

ambient temperature is reduced because radiant heat is not depth and 10 cm diameter and each large trap had depth of limited, however ambient temperature is influenced by 62.5 cm and 20 cm diameter. To distribute pitfall traps equally vegetation density, soil texture and rainfall [14]. across the one hectare we divided each plot into 25 subplots of In this study, we examined lizard species diversity and its 20 × 20 m each. The traps were placed at random within each association with major habitat features such as elevation, subplot while one large and two small traps were allocated to vegetation and soil structure through pitfall trapping in Qom each subplot and were buried into the ground so that their Province in Central Plateau of Iran which is considered as an entrances were flush with the ground surface. Pitfalls were arid environment. Species from the Palearctic, Oriental, and checked once every 7 days for 3 months. Afrotropical provenances [15,16] are found in this part of the Captured lizard species were identified using the field guide Central Plateau of Iran. However, little is known about the “The Lizards of Iran” [16]. Lizards were also sexed, weighed, lizard biodiversity in this region. We address three questions in and marked with a small dot of paint on its dorsal surface, hind this paper. 1) Does elevation influence lizard species diversity? and front leg, and tail, and then released [18]. If the lizard was 2) Are there additional habitat factors that play an important captured only once it had one mark, if it was caught twice it had role in species diversity? 3) What is the effect of habitat factors two marks and so on. on individual species abundances and how strong is their Vegetation sampling was conducted following the Braun- effect? Blanquet cover scale [19]. The vegetation quadrats were selected based on minimal area. In this method we established Materials and Methods a starting point for quadrat sampling. Point selection was based on the homogeneity of the vegetation in the studied quadrat. Study area The quadrat size was stratified to the vegetation type: annual, The Qom Province (34° 08’ to 35° 09’N, 50°08’ to 52°08’E, perennial, and shrub corresponded to quadrat sizes was datum = WGS 1984) forms the Northwest of the Iranian Central between 4 and 32 m2. We recorded the dominant and Plateau. It expands over a total area of approximately 11,238 companion species and their percent ground cover in each km2 and ranges in elevation from 600 to 3300 m. It has hot plot. We identified vegetation types for each quadrat based on summers and cold winters; average monthly temperatures in the presence of perennial species and also accompanying this area range from 15.2 to 40.1°C during summer and –1.7 to companion species. We named each vegetation type after the 18°C during winter. The average annual precipitation is 150 most dominant plant species. We then extrapolated the mm (range = 0.5 - 28.1 mm / month). The Qom Province measured percentages of vegetation ground cover of the ranges from a mountainous landscape in the south and west quadrats to the size of the one hectare plot in order to evaluate which includes the Zagros Mountain to a desert landscape in the abundance-dominance [19]. We summarized the total area Central Kavir in the east (Figure 1). each vegetation class (annual, perennial, shrubs) covered and divided it by the total area surveyed in each plot. We then Ethics further divided these relative values into 5 categories of relative The research was carried out under scientific permits from ground cover 1 (0–5%), 2 (6–25%), 3 (26–50%), 4 (51–75%) the Iranian Department of Environment and adhered to the and 5 (76–100%). We conducted vegetation surveys every welfare regulations in Iran. Under these regulations a month from February through October (spring through the end particular animal welfare license was not required, because we of summer) of 2006. We used overall plant cover, annual, did not remove, toe clip or sacrifice . All caught lizards perennial and shrub plant cover for each plot as vegetation were released at the point of capture after identifying the variables in our study. species. To determine whether soil structure impacts lizard habitat preference, we took a 0.5 kg core sample of soil from the Lizards and data collection center of each plot. We then sent these samples to the soil We conducted our study from June–August, 2006, in 12 one laboratory at the Pishahang institute, Qom Province, Iran. The ha plots. The plots were chosen such that they represented the soil texture was determined using the USDA textural system main elevation and vegetation gradients [8,17]. We used (clay, sand and silt) [20,21],water absorption was also satellite maps and field work to determine the elevation and measured [22]. Each of these soil factors were categorised to vegetation gradients and to assign plot locations (Figure 1). three levels, low (0-20%), medium (21-40%) and high (>41%). The field work included identifying agricultural lands, checking areas that were ambiguous in the map and gathering general Data analysis information of these areas in order to assist with selecting the In our statistical analysis of what drives lizard abundance plot locations. In our analyses we divided elevation into three and lizard diversity we used 10 environmental variables categories, low (less than 1100 m, five plots out of 12), consisting of elevation, 4 vegetation variables and 5 soil intermediate (between 1100 and 1800 m, three plots), and high structure variables. The dependent variable (abundance and elevation (above 1800 m, four plots). The measured elevations diversity indices) was calculated for each of 12 plots. The total of each plot are given in Table 1. Our sampling method number of each species caught in each plot over the sampling involved pitfall trapping. We placed 75 pitfalls (50 small and 25 period was then used for identifying species diversity and each large) in each one hectare plot. Each small trap had a 62.5 cm species abundance in each plot. We excluded 12 recaptured

PLOS ONE | www.plosone.org 2 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

Figure 1. The plots distribution in the Qom Province of Iran. The plot numbers correspond to those in Table 1. doi: 10.1371/journal.pone.0083890.g001 lizards from the analysis, to avoid pseudo replication. All other is an important step prior to calculating the Hill’s N2 index. individuals were caught only once during the study. Then we used regression with quadratic term to address how elevation affects species diversity. We used species diversity Relationship of species diversity and elevation (Hill's N2) as dependent factor and elevation as independent First we used Hill's N2 index (reciprocal of Simpson's index) factor. to determine species diversity for each plot [23]. We used this index because it is particularly suitable when a single species Relationship of species diversity (Hill's N2 index) and dominates the diversity in a plot, [24], which is the case in habitat variables using GLM some of our plots. We used the rarefaction method in the We used a generalized linear model (GLM) to examine the EstimateS freeware [25] to achieve equal sample sizes which effect of elevation, vegetation, and soil structures on species

PLOS ONE | www.plosone.org 3 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

Table 1. Number of individuals of each lizard species captured in each 1-ha plot (rows) in the Qom Province of Iran, and the species diversity among 12 plots (columns).

Plot number Species P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 (1452 m) (2942 m) (2390 m) (1931 m) (836 m) (2214 m) (951 m) (1069 m) (1289 m) (1709 m) (1041 m) (831 m) Ophisops elegans 23 - 1 5 - 3 1 - 8 95 - - Agamura persica 3 - - 1 1 - 9 1 - - 1 1 Bunopus carssicauda 6 - - - 6 - - 1 1 1 2 2 Trapelus agilis 16 - - - 5 - 13 14 4 14 22 - Trapelus lessonae - - - - - 1 ------Varanus griseus caspius ------1 - - - Eumeces schneiderii princeps - - - 1 - 1 2 4 8 - 25 - Laudakia caucasia - - 1 ------Phrynocephalus scutellatus ------4 - - 1 - - Phrynocephalus maculatus ------8 Mesalina watsonana 2 - - - - - 1 1 - - 7 - Eremias fasciata - - - - 18 ------Eremias velox - - - - 1 ------Eremias persica 8 - - 5 ------1 - Eremias andersonii - - - - 2 ------Number of specimens captured 58 0 2 12 33 5 30 21 22 111 58 11 Species diversity 3.93 0 2.5 2.7 3.6 3.01 3.5 3.02 4.15 2.47 3.2 2.4 ( Hill's N2 index) The values for species diversity derived from Hill's N2 index, as described in the text. doi: 10.1371/journal.pone.0083890.t001 diversity to address which habitat variables affect species matrix [28]. We conducted this analysis on 6 species (Table 2) diversity. We used the Hill's N2 index as the dependent after omitting all species that were only captured in one plot variable and three elevation categories, five overall plant cover and in low numbers or that were only recorded once. Before we categories and two soil water absorption categories as started our analysis we tested all habitat variables against each predictor factors in our final model. In order to identify all species abundance separately to find possible covariates as it significant interactions between habitat factors and species was described above. Then we analysed each species diversity (Hill's N2 index), we used a factorial model. Due to the separately with the species abundance as the dependent limited number of degrees of freedom (12 plots were sampled) variable, habitat variables which were not covariates were we could not include all sampled 10 environmental variables independent variables, and habitat variables which were into our model. Instead we followed Shenbrot and Krasnov [8] and used Pearson correlations to determine whether variables covariates as covariates. This analysis determines which were included as factors or covariates into the model. Variables habitat variable is the most important for each lizard species. that were highly correlated with the dependent variable (Hills Finally we used Ivlev's index as a measure of habitat selectivity N2 index; r>0.7) were included as covariates (annual and shrub [8]. This index ranked the preference of each lizard between -1 plant cover, percent of sand, silt and gravel) [26]. Finally, we and 1 for each habitat variable that we identified in the analysis calculated different models of GLM based on different of deviance. For this ranking we needed to divide each habitat combination of predictor factors that we added or excluded variable into intervals. We choose 5 intervals for each of our from model and selected the model with the lowest value of the habitat variables. These habitat intervals include elevation Akaike Information Criterion (AIC). This model shows the best (<1050, 1050-1550, 1551-2050, 2051-2551 and above 2551 possible relationship between habitat variables and species m), overall plant density (0%, 1-20%, 20.1-30%, 30.1-40% and diversity. The model was also ranked by chi-square to show 40.1%>), annual plant density (0%, <10%, 10.1-20%, 20.1-30% precision of the estimate of our model. and 30.1%>), clay (0%, <13%, 13.1-15%, 15.1-17%, and 17.1%>), soil water absorption <23%, 23.1-26%, 26.1-29%, Relationship of each species abundance and habitat 29.1-32%, 32.1>) and perennial plant density (0%, 50-60%, variables 60.1-71%, 71.1-81%, 81.1%>). To address which habitat factors affect individual species All analyses were conducted using R [29] and IBM SPSS abundances we used analysis of deviance assuming Poisson Statistics 20. error distribution [27], a method suggested for data with low numbers of caught individuals or with many zeros in the data

PLOS ONE | www.plosone.org 4 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

Table 2. Relationships between habitat features and lizard Table 3. Relationship of elevation and species diversity abundance in the Qom Province of Iran. according to regression analysis with quadratic term.

Parameter Sig. Coefficient t p value R2 Species Variable estimate F value df <0.05 AIC a 0.00400 2.742 0.023 0.770 Overall plant B -0.00002 -3.514 0.006 Ophisops elegans 0.170 6.53 1 0.030 253.00 cover doi: 10.1371/journal.pone.0083890.t003 Elevation 0.010 1.78 1 0.230 301.00 Gravel -0.050 0.66 1 0.440 410.00 captures) and Trapelus agilis (24.2%) were the most abundant Relative annual 0.030 0.57 1 0.470 413.00 lizard species in our plots. plant cover Bunopus Relative annual -0.090 7.39 1 0.020 40.41 Relationship of species diversity and elevation crassicaudatus plant cover Species diversity (Hill's N2) in each plot ranged from 0 to S.W. -0.180 0.33 1 0.590 44.00 4.15 (Table 1). The regression with quadratic term absorption Elevation -0.002 3.84 1 0.090 45.00 demonstrated that species diversity was highest at Relative shrub intermediate elevation (Table 3: Figure 2). 0.050 2.38 1 0.150 48.00 plant cover S.W. Relationship of species diversity (Hill's N2 index) and Trapelus agilis -0.270 14.04 1 0.005 69.88 absorption habitat variables using GLM Elevation -0.002 6.76 1 0.030 85.75 Species diversity (Hill's N2) in each plot ranged from 0 to Overall plant 4.15 (Table 1). The generalized linear model also 0.030 0.65 1 0.440 85.70 cover demonstrated that elevation was one of the most important Gravel 0.030 1.15 1 0.310 137.00 variable associated with species diversity among the 12 plots Eumeces (Table 4). The following interactions also had a significant schneiderii Clay -0.550 13.90 1 0.007 41.93 influence on the species diversity (Table 4): soil water princeps absorption × elevation, overall plant cover × elevation, and S.W. overall plant cover × soil water absorption × elevation. The -0.370 7.93 1 0.020 70.89 absorption lizard species diversity increased as overall plant cover and Elevation -0.003 5.84 1 0.040 84.00 elevation increased; however, species diversity declined as Relative annual elevation and overall plant cover were still increasing (Figure 0.030 0.64 1 0.440 124.00 plant cover 3). No lizards were found at elevations exceeding 3000 m (Plot Phrynocephalus Relative annual 2). Species diversity increased at low elevations (less than 0.100 4.83 1 0.005 22.64 scutellatus plant cover 1500 m), where overall plant cover reached moderate Elevation -0.002 1.84 1 0.200 26.00 percentages (25%–40%), and decreased at high elevations Overall plant (above 1500 m.), where overall plant cover rose to high -0.030 0.71 1 0.420 28.00 cover percentages (over 40%). When we included the next Gravel 0.001 0.02 1 0.890 29.00 interaction (soil water absorption), the trend of species diversity Mesalina S.W. remained the same as Figure 2, but species diversity was -0.680 7.95 1 0.030 23.29 watsonana absorption highest at low values of soil water absorption (less than 25 %), Elevation -0.004 16.53 1 0.004 26.00 with the moderate percentage of overall plant cover and low Relative elevation. Species diversity started to decrease when soil water perennial plant 0.100 7.49 1 0.030 26.00 absorption reached a moderate percentage (25%–40%), and cover decreased from mid to high elevations, from medium to high Overall plant 0.070 4.12 1 0.080 26.00 plant covers, and from low to medium percentages of soil water cover absorption (Figure 4). The models are ranked by AIC values, with the lowest values indicating the most important models. Only the species that had significant associations with habitat Relationship of each species abundance and habitat variables are shown, S.W. absorption = Soil water absorption; italics indicate that a variables variable was included as covariate; bold numbers indicate significant associations Analysis of deviance (Poisson error distribution) showed between variable and the abundance of that lizard species. statistically significant influences of habitat variables on species doi: 10.1371/journal.pone.0083890.t002 abundance in 6 species, although the significant habitat variables differed among species (Table 2). Elevation, overall Results plant cover, annual plant cover and perennial plant cover and soil water absorption were the most important habitat variables In total, we captured 363 individuals of 15 lizard species in that influenced lizard abundance. The overall patterns indicate the 12 plots (Table 1). Ophisops elegans (37.5% of all one or more of these habitat factors significantly influenced the

PLOS ONE | www.plosone.org 5 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

Figure 2. Pattern of lizards species diversity (Hill's N2 index) and elevation in the Qom Province of Iran. doi: 10.1371/journal.pone.0083890.g002 lizard abundance among different plots. For instance the Table 4. Relative importance of habitat variables and their abundance of O. elegans was significantly related to the overall interactions (×) to species diversity according to GLM plant cover, while elevation and soil water absorption were the (generalized linear model) analyses. main factors that influenced the abundance of T. agilis (Table 2). As we predicted, different habitat variables can influence the abundance of different lizard species, habitat selectivity Dependent Wald Chi- (Ivlev's index) also varied among species (Figure 5). By variable Independent variable Square df P value indexing the habitat selectivity for each species, we determined Species diversity Elevation 9.11 1 0.003 the species preference according to their related habitat (Hill’s N2) variables. Ophisops elegans avoided habitats with low overall Overall plant cover × Elevation 6.86 1 0.009 plant cover (Figure 5a). Bunopus crassicauda preferred Soil water absorption × Elevation 8.65 1 0.003 habitats with low annual plant cover (Figure 5b). Trapelus agilis Overall plant cover × Soil water 5.04 1 0.030 avoided high elevations and mountain habitats (Figure 5c), and absorption × Elevation soils with moderate and high-water absorptions (Figure 5d). Soil water absorption 3.68 1 0.060 Eumeces schneiderii princeps preferred moderate elevations Overall plant cover 1.16 1 0.280 Overall plant cover × Soil water (Figure 5e) and avoided soils with low and high percentages of 0.35 1 0.560 clay (Figure 5f) and moderate soil water absorption (Figure 5g), absorption The factors are ranked by chi-square values. but these preferences were not very clear. Phrynocephalus doi: 10.1371/journal.pone.0083890.t004 scutellatus preferred habitats with a moderate annual plant cover (Figure 5h). Mesalina watsonana seemed to prefer

PLOS ONE | www.plosone.org 6 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

Figure 3. Pattern of lizard species diversity (Hill's N2 index) in relation to overall plant density (percentage with star) and elevation in the Qom Province of Iran. Low elevation <1100 m, medium elevation 1101-1800 and high elevation >1801 m. doi: 10.1371/journal.pone.0083890.g003 habitats with moderate elevations (Figure 5j) and moderate soil diversity of lizards declines with increasing elevation, as is also water absorption (Figure 5k), but we could not identify a clear the case in birds and mammals [7]. Lizards often show high preference in this species according to perennial plant cover species diversity at intermediate elevations [17,31-33], but (Figure 5l). Phrynocephalus maculatus strongly preferred Site lower species diversity with increasing elevation. However, 12, which was a saline desert with low elevation and no McCain (2010) pointed out that temperature is the most vegetation. In addition, we captured one specimen of L. important habitat factor that correlates with these trends in caucasia in a mountain habitat (Plot 6). However, this species species diversity across elevations. In this study, the highest is a rock specialist and we were unable to place pitfall traps in species diversity occurred at an intermediate elevation (1289 rocky terrain. m) with a 25% overall plant cover. Above and below that elevation, species diversity decreased. Hence, our results from Discussion 12 plots at different elevation gradients support the concept that species diversity peaks at mid elevations (e.g., in our Plot The relationship between elevation and species diversity is 11). an important one in ecology, and elevation gradients are often Although elevation has a strong influence on lizard species associated with changes in habitat variables and community diversity, habitat heterogeneity can alter the direct impact of structure [6,11,30]. The average elevation in the Central elevation on lizard species diversity. Tews et al. [34] indicate in Plateau of Iran is about 900 m, but elevation increases to near their review that 85% of studies from 1960 to 2003 found 3000 m in some areas, especially in the Qom Province. Our positive correlations between habitat heterogeneity and results support the view of Heatwole (1982), that the species species diversity. In this study, lizard communities showed

PLOS ONE | www.plosone.org 7 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

Figure 4. Pattern of lizard species diversity in relation to soil water absorption (S.W.A), overall plant density (percentage with star) and elevation in the Qom Province of Iran. Low soil water absorption (L.S.W.A) <24.1, medium soil water absorption (M.S.W.A) >24.1. Plant density and elevation categories are same as Figure 2. doi: 10.1371/journal.pone.0083890.g004 different patterns when habitat variables changed rapidly at the Processes of habitat selection in species involve different same elevation. At two low-elevation plots, the species factors. One important aspect of this process is habitat diversity was lower (2.4) in one plot (plot 12 at 831 m) than in variability [35,36]. As we have shown, habitat variables affect the other (3.6 for plot 5 at 836 m). A lack of vegetation and a species diversity. The process of habitat selection by each high percentage of salt and clay were associated with the species in our study sites explains how different habitat reduction in species diversity in this case. variables affect species abundance in each plot. In other The different impact of each interaction on species diversity words, response of lizard species to different habitat factors is a result of species adaptations and physiognomic conditions and the degree of their response to habitat factors (-1, 0, 1) is of the study area. For instance plots between 1000 and 2000 m one of the reason that overall species diversity is varied among did not show a clear pattern of species diversity according to sites, even if some of the habitat variables are almost equal. elevation alone, when we added other habitat factors such as The present study was designed to determine the effect of vegetation cover and soil water absorption, the pattern became habitat variables on lizard species diversity based on pitfall clear and showed the maximum species diversity in sites with low altitude, low soil water absorption and medium plant cover. trapping in Qom Province in Central plateau of Iran. This study One of the explanations of this change in species diversity is has found that generally elevation is one of the main that environmental variables have a different value for each environmental variables that changes species diversity. The species, and they can be species-specific, as we have shown second major finding was that extreme change in one habitat for 6 species in the second part of our results. variable can override the effect of another habitat variable and

PLOS ONE | www.plosone.org 8 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

Figure 5. Habitat variable ranks for six lizards species in the Qom Province of Iran that had significant relationships with habitat variables. The value of the selectivity index is shown and each variable is divided to five intervals. doi: 10.1371/journal.pone.0083890.g005 influence species diversity, such as species diversity in plot 12 Game wardens of Department of Environment of Qom and plot 5. Province especially Mr Naser. Prof. M. Bull and Prof S. C. This information can help us with future surveys to Anderson for their great advice. Dr A. L. Fenner, Dr S. T. Leu understand how these habitat changes could affect lizard and Mrs. J. Schofield for their suggestions and editing. Prof. H. species abundance and predict future changes in their Liaghati, Prof. Kiabi, Mr. M. Zohary, Mr. Alikhani and Mr. Kakai abundance. This has implications for conservation are thanked for their efforts to ensure that the vehicles and management that can help us to determine how lizard species other materials used in this demanding field work were abundance is affected by a change in environmental variables available, and also we would like to thank the volunteer due to human activity or climate change. students (Hadi Heidary, Meisam Habibi) for establishing the plots. Acknowledgements Author Contributions We wish to thank the Department of Environment of Qom Province and Environmental Sciences Research Institute Conceived and designed the experiments: ME FA HM ARM AA (ESRI) in Shehid Beheshti University for providing an ASM. Performed the experiments: ME FA HM ARM. Analyzed opportunity for our study, within the framework of the project of the data: ME. Contributed reagents/materials/analysis tools: aquatic and terrestrial habitats of Qom Province (127-9112). ME FA. Wrote the manuscript: ME ARM.

References

PLOS ONE | www.plosone.org 9 December 2013 | Volume 8 | Issue 12 | e83890 Ecological Associations of Lizards in Iran

1. Cody ML (1968) On the methods of resource subdivision in grassland 18. Brown WS, Kéry M, Hines JE (2007) Survival of Timber Rattlesnakes bird communities. American Naturalist 102: 107-147. doi: (Crotalus Horridus) Estimated by Capture–recapture Models in Relation 10.1086/282531. to Age, Sex, Color Morph, Time, and Birthplace. Copeia 2007: 656-671. 2. Huey RB, Pianka ER, Schoener TW (1983) Lizard ecology: Studies of a 19. Braun-Blanquet J (1964) Pflanzensoziologie: Grundzüge der model organism. USA: Harvard University Press. vegetationskunde. USA: Springer. 3. Perry G, Garland T (2002) Lizard home ranges revisited: Effects of sex, 20. Smith KA, Mullins CE (1991) Soil analysis: physical methods. USA: M. body size, diet, habitat, and phylogeny. Ecology 83: 1870-1885. Dekker: 690. Available online at: doi: 21. Brown RB (2003) Soil Texture. University of Florida. 10.1890/0012-9658(2002)083[1870:LHRREO]2.0.CO;2 22. Day PR (1954) Soil moisture tension measurements: Theoretical 4. Vitt LJ, Pianka ER (1994) Lizard ecology: Historical and experimental interpretation and practical application. Clays Clay Miner 3: 557-566. perspectives USA. Princeton University Press. doi:10.1346/CCMN.1954.0030144. 5. Stamps JA (1983) Sexual selection, sexual dimorphism, and 23. Hill MO (1973) Diversity and evenness: A unifying notation and Its territoriality. In: RB HueyER PiankaTW Schoener. Lizard ecology, consequences. Ecology 54: 427-432. doi:10.2307/1934352. studies of a model organism. USA: Harvard University Press. pp. 24. Kelly MG (2001) Use of similarity measures for quality control of 169-204. benthic diatom samples. Water Res 35: 2784-2788. doi:10.1016/ 6. Lomolino VM (2001) Elevation gradients of species-density: Historical S0043-1354(00)00554-6. PubMed: 11456180. and prospective views. Global Ecology & Biogeography 10: 3-13. doi: 25. Colwell RK (2000) EstimateS: statistical estimation of species richness 10.1046/j.1466-822x.2001.00229.x. and shared species from samples, version 8. Robert K Colwell. 7. Rahbek C (1995) The elevational gradient of species richness: A 26. Miyanishi K, Kamo Y, Ihara H, Naka T, Hirakawa M et al. (2006) Risk uniform pattern? Ecography 18: 200-205. doi:10.1111/j. factors for dysbaric osteonecrosis. Rheumatology (Oxford) 45: 1600-0587.1995.tb00341.x. 855-858. doi:10.1093/rheumatology/kel013. PubMed: 16436490. 8. Shenbrot GI, Krasnov B (1997) Habitat relationships of the lizard fauna 27. Baker RJ, Nelder JA (1978) The GLIM system release 3, numerical in the Ramon Erosion Cirque, Negev Highlands (Israel). J Zool 241: algorithms group. United Kingdom: Oxford. 429-440. doi:10.1111/j.1469-7998.1997.tb04835.x. 28. Towns DR, Elliott GP (1996) Effects of habitat structure on distribution 9. Tuomisto H (2010) A consistent terminology for quantifying species and abundance of lizards at Pukerua Bay, Wellington, New Zealand. N diversity? Yes, it does exist. Oecologia 164: 853-860. doi:10.1007/ Z J Ecol 20: 191-206. s00442-010-1812-0. PubMed: 20978798. 29. R Development Core Team (2005) R: A language and environment for 10. Gaston KJ, Spicer JI (2004) Biodiversity: An Introduction, 2nd ed. statistical computing, reference index version 2.0.1. Vienna, Austria: R Australia: Blackwell. Foundation for Statistical Computing. 11. Meik JM, Lawing AM (2008) Elevation gradients and lizard assemblage 30. McCain CM (2005) Elevation gradients in diversity of small mammals. structure in the Bonneville Basin, western USA. J Arid Environ 72: Ecology 86: 366-372. doi:10.1890/03-3147. 1193-1201. doi:10.1016/j.jaridenv.2008.01.003. 31. Kasangaki A, Kityo R, Kerbis J (2003) Diversity of rodents and shrews 12. Hofer U, Bersier LF, Borcard D (1999) Spatial organization of a along an elevational gradient in Bwindi Impenetrable National Park, herpetofauna on an elevational gradient revealed by Null model testes. south-western Uganda. Afr J Ecol 41: 115-123. doi:10.1046/j. Ecology 80: 976-988. Available online at: doi: 1365-2028.2003.00383.x. 10.1890/0012-9658(1999)080[0976:SOOAHO]2.0.CO;2 32. Li JS, Song YL, Zeng ZG (2003) Elevational gradients of small mammal 13. Heatwole H (1982) A review of structuring in herpetofaunal diversity on the northern slopes of Mt. Qilian, China. Global Ecology & assemblages. 1-1988 in Scott Jr. NJ. editor. In Hofer U, Bersier LF, and Biogeography 12: 449-460. doi:10.1046/j.1466-822X.2003.00052.x. Borcard D (1999) Spatial organization of a herpetofuna on an 33. Sanders JN, Moss J, Wagner D (2003) Patterns of ant species richness elevational gradient revealed by Null model tests. Ecology 1980: along elevational gradients in an arid ecosystem. Global Ecology & 14. McCain CM (2010) Global analysis of reptile elevational diversity. Biogeography 12: 93-102. doi:10.1046/j.1466-822X.2003.00324.x. Global Ecol Biogeogr 19: 541-553. 34. Tews J, Brose U, Grimm V, Tielbörger K, Wichmann MC et al. (2004) 15. Sindaco R, Jeremcenko JV (2008) The Reptiles of the Western Animal species diversity driven by habitat heterogeneity/diversity: the Palearctic. 1. Annotated Checklist and Distributional Atlas of the importance of keystone structures. J Biogeogr 31: 79-92. doi:10.1046/j. Turtles, Crocodiles, Amphisbaenians and Lizards of Europe, North 0305-0270.2003.00994.x. Africa, Middle East and Central Asia. Italy: Edizioni Belvedere. 35. Shenbrot GI (1988) On the estimation of factors affecting lizard 16. Anderson SC (1999) The lizards of Iran. USA: Society for the Study of distribution and abundance in the deserts of South Bukhara Region. Amphibians and Reptiles. Ekologiya (USSR) 5: 50-56 (in Russian, English translation) 17. Fischer J, Lindenmayer BD (2005) The sensitivity of lizards to 36. Shenbrot GI (1991) On the environmental factors affecting desert lizard elevation: A case study from south-eastern Australia. Diversity & distribution and abundance. General Meeting of Societas Hepetologica Distributions 11: 225-233. Europaea. Budapest. p 82.

PLOS ONE | www.plosone.org 10 December 2013 | Volume 8 | Issue 12 | e83890